The long-term goal of this research is to understand the neural basis for the perception of communication sounds in the cerebral cortex of primates and the fundamental neural mechanisms that sub serve cortical representations of these biologically important sounds. We have established an experimental model, the common marmoset (Callithrix jacchus), to address these issues. This model system provides several important advantages over other species, namely, a rich vocal repertoire;an auditory cortex that lies largely on the lateral surface of the cerebral cortex and an extremely high reproductive rate while in captivity. In the present application, we will continue this line of research along two fronts: neural representations of marmoset vocalizations and cortical coding mechanisms that are relevant to the processing of primate vocalizations. The emphasis of the present application is to elucidate functional differences and the transformation between the primary auditory cortex (A1) and the lateral belt areas (LB). In Specific Aim 1, we will quantitatively characterize spectral selectivity of neurons in A1 and LB using random spectrum stimuli (RSS) that we have recently developed. We will investigate temporal selectivity of LB neurons in Specific Aim 2. Our preliminary data indicated that, in addition to their spectral selectivity, LB neurons showed further enhanced responses if proper temporal modulations were introduced. These experiments are the natural continuation of our extensive work on the temporal selectivity of A1 neurons and will shed light on the hierarchical processing of time-varying signals in the auditory cortex of non-human primates. In Specific Aim 3, we will quantitatively define neural selectivity for marmoset vocalizations in A1. We will carry out these experiments using synthetic vocalizations based on the statistics of marmoset vocalizations obtained in the P.I.'s laboratory, which can be easily manipulated in both spectral and temporal domains. In Specific Aim 4, we will test the hypothesis that LB neurons have a greater selectivity for marmoset vocalizations than do A1 neurons. Findings of the present study will contribute to our basic understanding of the cortical representation of complex acoustic stimuli, and will have implications for the neural basis of human speech perception and for designing better hearing aids and) prosthetic devices for the deaf and hearing-impaired.